View
3
Download
0
Category
Preview:
Citation preview
Stylistic iconicity and Love's two orders oflanguageToolan, Michael
DOI:10.1016/j.langsci.2016.09.016
License:Creative Commons: Attribution-NonCommercial-NoDerivs (CC BY-NC-ND)
Document VersionPeer reviewed version
Citation for published version (Harvard):Toolan, M 2016, 'Stylistic iconicity and Love's two orders of language', Language Sciences.https://doi.org/10.1016/j.langsci.2016.09.016
Link to publication on Research at Birmingham portal
Publisher Rights Statement:Checked 13/12/2016
General rightsUnless a licence is specified above, all rights (including copyright and moral rights) in this document are retained by the authors and/or thecopyright holders. The express permission of the copyright holder must be obtained for any use of this material other than for purposespermitted by law.
•Users may freely distribute the URL that is used to identify this publication.•Users may download and/or print one copy of the publication from the University of Birmingham research portal for the purpose of privatestudy or non-commercial research.•User may use extracts from the document in line with the concept of ‘fair dealing’ under the Copyright, Designs and Patents Act 1988 (?)•Users may not further distribute the material nor use it for the purposes of commercial gain.
Where a licence is displayed above, please note the terms and conditions of the licence govern your use of this document.
When citing, please reference the published version.
Take down policyWhile the University of Birmingham exercises care and attention in making items available there are rare occasions when an item has beenuploaded in error or has been deemed to be commercially or otherwise sensitive.
If you believe that this is the case for this document, please contact UBIRA@lists.bham.ac.uk providing details and we will remove access tothe work immediately and investigate.
Download date: 01. Feb. 2019
brought to you by COREView metadata, citation and similar papers at core.ac.uk
provided by University of Birmingham Research Portal
1
Stylistic iconicity and Love's two orders of language
Michael Toolan
My aim here is to put integrationism’s anti-systematicity into close conjunction with
stylistics’ craving for linguistic explanation of literary effects, to see whether any kind of
rapprochement, even of conceptual inclusion of the latter within the former, is remotely
possible—or whether such an effort only highlights the mutual incompatibility of the two
mind-sets. Love’s idea of two orders of language activity helps to focus the difficulties.
Applying these to the practice and principles of stylistics, they prompt several basic
questions: Is stylistics doomed to the adoption of one ‘second order’ descriptive machine
after another, in a futile effort to capture the essence, the life, of the literary texts it analyses?
Are readers’ actual engagements with poems, their ‘first order’ experiencing of poems,
never to be satisfactorily or usefully explained by these abstracting, unit-manipulating,
second-order systems? In moving towards answers to these questions I will suggest that
stylisticians are not unaware of their difficulties (sometimes more aware than other linguistic
analysts of texts), and that one concept that some have adopted as part of an effort to forge a
strong connection from the second-order description back to the first-order experience, is
that of iconicity. I discuss whether ‘iconicity’ is stylistics’ attempt to bridge the second-
order/first-order gulf.
1. Integrationism and Literature
Few integrationists have much to say about literature, perhaps regarding it as a specialist
interest of the educationally privileged in modern societies—writers and readers-- whose
outpourings have already received a disproportionate amount of linguists’ attention.
Certainly the 1066-and-all-that history of linguistics describes European philologers
devoting much energy to linguistic analysis of each language’s literary canon studied
diachronically, and minimal attention to contemporary speech and its synchronic
systematicity. Then came Saussure and Boas, the revaluation of now mechanically-
recordable speech, and a general turning away from regarding language-study and linguistics
as having much centrally to do with literature, or vice versa. After all, everybody uses
language every waking hour in all sorts of complex ways, and many do so without reading a
word of literature, so the latter is clearly a linguistically peripheral cultural practice.
That, to my mind, is only part of the story. Another part would argue that Literature
is modernity’s religion of the book (it also often provides a good test for linguists’ theories).
For a significant proportion of literate people in western societies especially, and since the
early 19th
century mostly, spiritual guidance, self-understanding, explanation of what makes
for a thoughtful, compassionate, ethical co-existence with all other creatures on the planet,
has come as much, if not more, from their literary reading than from their reading of the
Bible, the Torah, the Koran, or other holy scriptures. It is not clear to me that literature is
still, in the 21st century, the primary spiritual guide for modernity’s adherents—arguably
songs to some extent and films especially, in various forms, have taken over the primary
position. But by films here I mean narrative films--films that tell stories and almost
invariably include language in the telling, most prominently in characters’ speech.
Narratives that use language (many do not) are arguably the master form, when people look
for guidance and understanding. Notwithstanding the richness and subtlety of ideas that
language-free film or music or visual art can prompt in recipients, literature retains its own
2
distinct position, by virtue of the complexity of the articulated situations it can convey
through language.
From the era of structural linguistics onwards, most academic linguists in the West
have had little or no interest in literature (with a few notable and not-entirely-mainstream
exceptions, such as Jakobson, Kiparsky, Halliday, and Hymes), on the grounds that literature
is remote from ‘ordinary language’, the canonical speech situation, and the mysteries of
language acquisition by infants. The art in literary language, and literary culture as a whole,
was implicitly judged by mainstream linguistics to be ‘artificial’, cultural rather than natural,
of doubtful authenticity, often ‘contrived’, and absolutely the wrong place to look to observe
the workings of ordinary language in its basic functions. If linguists analysed poems at all,
there was a tendency to emphasize that each poem was sui generis, and related obliquely to
the rules and units of ordinary language; for a while some generative linguists like Thorne
proposed that each new poem had its own distinct language and grammar. But among those
linguists who did explore poetics, such as Jakobson and Thorne, iconicity is often an at least
implicit criterion of explanation.
Integrational linguists may have neglected literary texts just as much as mainstream
linguistics did, but for a different reason: how can there be any great ‘conveyed complexity’
in literary texts, worthy of analysis, if in line with the freedom of the sign-making individual
that integrational semiology asserts, every different recipient is entitled to their own unique
understanding of any given text, any purportedly artistic communicative act? But that
theoretical principle applies to all texts and communicative events, and not exceptionally to
poems. Nor should it be treated as licensing widespread communicational
misunderstanding, but a more nuanced model of communicational accord, involving
context-shaped coordination of words and acts. No-one writes a book like the Language
Myth, urging a demythologized and redefined linguistics, with the expectation that every
reader will make sense of it so differently from their fellow-readers that all mutual
understanding is lost—least of all Roy Harris.
Integrational linguistics could have been among the last to accept the exiling of
literature from the attention of linguists, since literature is arguably a best exemplar of its
central claims. Where two or three gather together to read and discuss their reactions to a
poem that is new to them, the experience will be palpably remote from what the language
myth promises: no smooth mind-to-mind transfer of ideas, no rapid and uniform matching of
forms to meanings, and no great certainty as to an agreed understanding at all.
Integrational semiology says all meaning, all sign-making and sign-interpreting is
‘now’ (not necessarily the same ‘now’ for all parties: I can leave some sticks on the beach in
a particular configuration one evening, with a view to you seeing and ‘reading’ them the next
morning). In a sense, with regard to meaning and communication, integrationists are always
saying ‘you had to have been there’ (to ‘get’ what was signified by and for the parties
involved).
On the other hand there is something we must set beside or against readers’
integrationism-warranted diversity of interpretations (sometimes an unmanageable
diversity—but then, why value ‘manageability’?): it is the fact that poets and stylisticians are
no less immune to seduction by the language myth of fixed-code telementation than anyone
else. I suspect that we stylisticians, negotiating a kind of repression, intermittently simply
know that linguistic communication—a poem, say—cannot possibly be based upon a fixed
code of form-meaning pairs known or knowable by all. But craving the permanent and
enduring, something that transcends the ephemeral moment and the never-to-be-recovered
individual context, we contrive to explain what we see as the main enduring or promising-to-
endure meanings and significance of specific linguistic communications (poems, and extracts
from novels and plays), on the grounds that these are one kind of consolation—defiance
3
even—of the variability, indeterminacy, entropy, dissolving, and death, which we also know
to be the case. And in that yearning for the permanent text, the promise made by the idea of
iconicity in literature is seductive. It proposes that, for these particular language events, there
is no specific ‘there’: as long as you are English-proficient, you can read a Shakespeare
sonnet at any time, in any place, and some of the experience or meaning of the poem will (or
should) be yours, whatever the contextual circumstances, because iconicity claims a natural
fit between form and meaning. The inescapable test case, especially in the 400th
anniversary
of his death, is Shakespeare--for example, his sonnet #55:
Not marble, nor the gilded monuments
Of princes, shall outlive this powerful rhyme;
But you shall shine more bright in these contents
Than unswept stone besmear'd with sluttish time.
When wasteful war shall statues overturn,
And broils root out the work of masonry,
Nor Mars his sword nor war's quick fire shall burn
The living record of your memory.
'Gainst death and all-oblivious enmity
Shall you pace forth; your praise shall still find room
Even in the eyes of all posterity
That wear this world out to the ending doom.
So, till the judgment that yourself arise,
You live in this, and dwell in lovers' eyes.
2. Love’s two orders of language activity
Nigel Love’s 1990 distinction between first order and second order language activity excited
me when I was writing Total Speech (1996), and it is no less stimulating an idea to me now.
He begins (1990: 96) with Bloomfield’s “fundamental assumption of linguistics” (1935: 144),
that for some speech community or other, some utterances are alike as to form and meaning,
thus allowing the beginnings of a system of repeated signs to be thinkable. Without such
recurrent ‘sames’, ordinary linguistics would not seem possible at all. But the assumption
itself is magic wand-waving by the theoretician, Love suggests, and Bloomfield made no
attempt to provide evidential support for it—instead by implication he treats orthographic
writing as the terra firma justifying the assumption. If I write Odense is rainy and you write
Odense is rainy these are said to be alike as to form and meaning, notwithstanding
differences of handwriting, ink, surface, what have you. How do we know? Because they
are written the same, and anyone who reads one aloud will make much the same sounds as
when reading the other aloud. In short, the circularity or question-begging is (incompletely)
masked by writing, or what writing seems to entail.
Love further suggests that the making of a distinction between types and tokens is a
necessary condition of ‘generalizing discourse’ about anything, and that “this capacity of
language to facilitate generalizing talk about the world is unremarkable” (97). What he
intends here—I believe—is not any claim about the ‘necessity’ of a philosophical explanation
in terms of tokens and types, but about the interdependence, without priority, of several ideas:
perceptions of sameness; performing a sameness (e.g. saying or writing what you regard as
the same thing again); abstraction; and generalization. Any one of these entails the others.
But we have no grounds for assuming that our abstraction, our sorting of instances into types,
is itself identical to that of others. We use this capacity to treat utterances as if they were
4
recurrent variant instances of underlying invariants all the time in linguistic communication.
The mistake, Love suggests, is to assume uniformity among all speakers in a speech
community, as to what count as instances and underlying invariants. At several points Love
argues that the problem is not with the idea of abstraction itself, of individual communicators
devising a cognitive categorization of the utterances they encounter or produce. The problem
is with any assumption that this abstracting operates on a unified, determinate, homogeneous
substrate: “The problem is not the idea of abstractions underlying speech, but the idea of
determinate systems of abstraction underlying speech” (Love 1990: 108) The world is not the
same for everyone, not has ever been.
First-order linguistic communication is the use of language to communicate
(informing that use with your past linguistic, indeed interactional, experience or history,
mostly spontaneously and unreflectingly). It is a ‘doing’; it can be minimally self-conscious,
and is often so in the very young, those suffering from dementia, the rest of us in moments of
emotional extremity, and perhaps, sometimes, in self-communication. Second-order linguistic
communication comprises every kind of articulated reflection upon first-order linguistic
communication, and particularly the verbal parts of it. It assumes and exploits self-
consciousness, reflexivity, our powers of memory and our ability to treat temporally
discontinuous material as repetitions and parts of a pattern. It also assumes our powers of
imaginative projection: we can, vividly enough, imagine ourselves stepping out of the on-
flowing river of life to appear to stop living now, while we ponder and report on what we or
someone else said then. So second-order linguistic communication is metalinguistic activity
of all kinds. But is this metalinguistic activity exclusively second-order? No. I cannot deny
that even in maximally ‘engaged’, minimally reflexive first-order language-using activity, we
still find speakers using kinds of metalanguage, albeit more evaluatively than analytically
(e.g. when they say someone is a smooth talker or uses filthy language or sounds angry all the
time).
I therefore do not find the ‘two orders’ distinction unproblematic. In particular it is
quite hard to recognize a ‘first-order’ use of language that is utterly devoid of the proposed
defining features of second-order language communication (and, perhaps, vice versa too).
Thus in a recent commentary on the distinction sent to all participants prior to the Odense
Symposium, Love says that in first-order communication the speaker has to creatively deploy
parts of their past experience to maximise likelihood of communicative success, and the
hearer is supposed to ‘do their best’ to understand the speaker. But for a hearer to ‘do their
best’ seems to imply some comparative self-assessment on their part, a comparing of their
current effort with alternative qualities of effort, which looks very much like a second-order
activity. Likewise the speaker’s task of selecting clues from past experience also suggests a
kind of second-order detached reflection upon both present and past communicational
activity.
While sometimes (or even always) difficult to maintain, the ‘two orders’ distinction is
also useful—a usefully extreme encapsulation of two quite different ways of thinking about
and using language, which might be summarized, again in extremis, as the ordinary person’s
approach to language (full integration) and the academic linguistician’s approach to it (at
least partial segregation). At the very least, Love is arguing that we need to be attentive to
the differences between the two orders, and to beware of presuming that a commentary
couched in the terms of the second-order will necessarily yield insights into the workings of
linguistic communication in the first-order. I am assuming also that a redefined linguistics of
an integrationist or related kind would hope to clarify the relations between the second and
first orders, and make second-order academic linguists more responsive to, more engaged
with, first-order activity. Whether that is a matter of building bridges between the two, or
contriving a merger (an absolute separation would be a policy of despair, presumably) is less
5
clear to me.
Some of the thinking that developed into the idea of two orders of language activity
are detectable, I believe, in an earlier paper critiquing the fixed-code theory (Love 1985).
There, Love uses the variant pronunciations of postmen, the philosophers’ complicated
attempts to explain the allegedly paradoxical nature of negative existential sentences, and
variable informant responses about the propositional entailments of sentences like my son
threw a brick at the window, to make the point that these are linguists’ language-games where
various moves (objections, reformulations) are allowed but where there is a limit, a point
where you are deemed no longer to be playing the game. At that point you are ‘cheating’
(Palmer), beyond the pale, not doing linguistics. Such ‘cheating’ might prompt the rhetorical
question Is nothing sacred? (with its own negative existential)—because in the language
games of orthodox linguistics something certainly is sacred, and that is the commitment to
the idea of unambiguous form-meaning pairings as the foundation of a language system.
Holding that principle sacrosanct, the linguist game-player will have no time for Love’s
observations with regard to the variability of propositional entailments depending on the
communicational needs on a particular occasion of use. Implicit in orthodox linguistics’
disallowing of ‘beyond-the-pale examples’, I believe, is a sharply circumscribed conception
of first-order language activity, upon which (and only upon which), linguists’ sophisticated
second-order analytical machinery can reliably work. The segregational linguist will dismiss,
as ‘cheating’, the idea that “the relation between propositions and the forms of words that
express them is not such that the former can be read off directly, in the absence of any
context of utterance, from the latter” (1985: 3). Because the orthodox linguist is so firmly
committed to the describing of a fixed code pairing of abstraction conceived of as forms and
meanings (for thought-communication) “our inability to say precisely what might be
communicated by my son threw a brick at the window is taken as no hindrance to the
enterprise of determining, in the abstract, the meaning of that expression” (1985: 5-6).
Actual communicational practices using language are held at some distance from this
idealising and arguably misrepresenting endeavour.
One of the important things that Love’s first order/second order naming achieves is to
make clear that the systematisation of verbal communication into something conceived of as
a language with a set of items and rules for their combination (what I’ve called the
codification impulse) comes after, or second. Orthodox linguistics has managed to induce
many to believe that the order is exactly the reverse, and the terminology of system and
instantiation aids in that misrepresentation, as do the many game and dance analogies. First
you have chess with all its rules, and then you have a rich ‘creative’ variety of possible legal
moves that follow those rules (and if you devise a board game with 36 squares and no knights
then you’re not playing chess at all, so, again, beyond the pale).
Love suggests that perhaps the most primitive manifestation of our decontextualising
ability, our ability to move to the second order of language activity, is displayed by our
capacity to respond to the metalinguistic question “What did you say?”—or, I would add,
“What did they say?”. Any non-literate person who asks either of these questions has a
concept of ‘sayings’, of sequences of oral sounds understood to play an important role in the
expression of currently relevant meanings, sequences sufficiently object-like that speaker and
addressee can try to orient to them and inspect them. Love pinpoints the metalinguistic
question as evidence of the language-user’s grasp of the possibility of contextually-relevant
repetition. For me, it is equally important as evidence of a kind of intersubjective faith: a
trust that it is possible, to all important intents and purposes, for someone to replay now and
here what someone said far away and long ago. In this replaying, this reported speech, there
is no identifying of a linguistic unit, let alone an identical linguistic unit. In reported speech
something “relevantly similar”, to use Love’s formulation, is surely all that interactants who
6
trust in the possibility of it expect. Elsewhere (1990: 98), Love suggests that when A asks for
an exact repetition from B, the judgement that B’s second effort is an exact repeat is
grounded in no transcending abstract scheme, no autonomous system, nothing beyond
situation-internal conformity with the idea that it is, in all the interactional circumstances, a
valid repeat. Even to me, carrying the integrationist virus (Owen nnnn), this kind of non-
explanation feels like the sort of hand-waving that most linguists find exasperating; but
laying claim to any more determinate basis for sorting the repetitions from the variations is
arguably to descend into scientism. Besides, it is surely no accident that only in legal and
other quarrelsome situations that ideas of verbatim reporting, of X saying ‘exactly’ what Y
said, come to the fore; and even these never really cease being a kind of loose talk (to use
relevance theory’s useful concept), a context-specific acceptance that such and such will
count as ‘an exact repetition’ of some previous formulation. The looseness begins with the
indeterminacy as to what verbatim or faithful themselves mean.
“What did you just say?” and “What did they say?”, and the possibility of giving
answers to such questions, encapsulate metalinguistics, or second-order language activity and
thinking, in the most compelling way. They invite us to imagine a language-using world
where such questions are missing. (They are absent from Wittgenstein’s four-term builder’s
language, of Block!, Slab! etc.,--but that is an implausibly reduced language.) I cannot
imagine that such a world would not also lack questions like “‘What happened?” and “What
did you just do?”, so “What did you say?” is an integral aspect of our capacity for partially
removing ourselves from the present flow of experience to the point that we temporarily
focus upon spatiotemporally removed activity, and put current activity in the background. It
is what deixis is all about (deixis suitably broadly conceived being, in my opinion, what the
final chapter of The Language Myth is also all about). We cannot, however, escape that
embeddedness in the spatiotemporal present, as Harris’s concept of cotemporality
emphasises.
Just as ‘what did you say’? is the canonical metalinguistic question, the very idea of
reported speech lies at the heart of linguistics; furthermore, in one form or another, it is a
generator of endless papers and theorising in narratology and cognitive poetics—e.g. Text
World Theory (see, e.g., Werth 1999, Gavins 2007), just one of several attempts to develop a
deictic or perspectival model of the way stories can be embedded within stories recursively,
giving rise to complex interpretive tasks for the reader, assessing the varying reliability of
narrators at different levels of embeddedness. In short and in essence, the problem of hearsay
evidence. You heard that X said Y but you didn’t yourself hear X say Y, did you? More
complex yet than the literary representation of bog standard direct speech and indirect speech
is that written discourse which has many of the formal characteristics of indirect speech (e.g.
past tense, 3rd
person pronouns) but when read aloud nevertheless ‘sounds’ like a direct
voicing of a character’s words or thoughts: a discourse presentational format that ‘cheats’,
going beyond the pale of grammatical direct and indirect discourse. No topic is dearer to the
hearts of narratologists and novel criticism than such Free Indirect Style (to give one of its
myriad names), a sophisticated literacy mutation of the reported speech genus, raising
difficult questions for the reader about quite whose judgements (the narrator’s or a
character’s?) are being presented, as from what time, and in what place.
In the stylistic analysis of literature, there is a powerful wish to describe or identify, if
possible, the abstract linguistic systems that underlie the poem or style under scrutiny and to
try to persuade others that they too, if they do enough second-order reflection upon their first
order experience of the poem, must agree that just these systems, these units, these forms,
have very much the meanings that the specific stylistician proposes. Or, making concessions
to different readers’ racial, sexual, etc identities, stylisticians will allow some divergences in
interpretation, but will tacitly cling to the idea of convergence, of a large and central area of
7
‘shared experience’ by virtue of stable forms triggering or containing stable meanings. The
alternative is too distasteful—it is the suggestion that what I take to be the forms and
meanings of Shakespeare’s sonnet 55 might be incalculably different from what the next
person takes that poem’s forms and meanings to be, that in fifty, ten, or five years’ time the
view might emerge that this sonnet is as inconsequential as yesterday’s free newspaper. To
stylisticians it makes no sense to imagine that in a hundred years time people might agree that
sonnet 55 is rubbish.
3. Taylor and the politics of language
In the same long ago Redefining Linguistics volume that carried Nigel Love’s paper appeared
also Talbot Taylor’s plea that a redefined linguistics pay proper attention to the kinds of
normativity in language that, from the perspective of the language-user, are so important.
Normativity, socio-political pressures to conform or converge in our usage, our form-
meaning pairings and our form- and meaning-selections, are a prominent part of the ordinary
socialised language-user’s understanding of what language is, and affects their own uses of
language and their evaluations of others’ uses. An integrationally redefined linguistics
should attend to this. A linguistics which claims to espouse pure descriptivism is failing to
engage with “the important issues of the politics of language”, Taylor says, adding:
From the perspective of the language-user, concepts such as ‘right’, ‘wrong’, ‘good’,
and ‘correct’ (no less than ‘word’, ‘meaning’, and ‘sentence’) are necessary to an
understanding of what a language is. (1990: 142, italics in original)
I would agree, but the evaluative judgements that are so important a part of understanding of
a language for a language-user do not stop there, with ideas about good, bad, correct and
incorrect. What about language that gets talked about because someone deems it beautiful, or
ugly, or forceful, or mysterious, or vivid, dynamic, transcendent, agitated first but calm later,
dreary, clotted, stodgy, paradoxical… in fact the whole panoply of forms of judgement that
readers normatively orient to when they feel impelled to talk about samples of language at
all?
It is possible that readers may not feel impelled to talk about any of the language that
they themselves produce or that they encounter coming from others, and likewise perhaps we
could go through life drinking various kinds of wine—the way bees suck nectar—and never
pass any kind of reflective judgement on them. But as Love has argued, it is hard to imagine
anything approximating a spoken language getting under way without a concomitant
reflexive (and often normative) set of practices of second order commentary also developing
in parallel. With writing it all gets much more developed and sophisticated (on this topic,
Teubert 2010 is an important recent contribution). With literature, commentaries on
literature, and those reversions to form- and meaning-analyses that claim to illuminate those
commentaries (aka stylistics), reflexive commentary is thoroughly entrenched. We could say
it is unnecessary to do these things—just drink the wine, just listen to the poem—but is there
any such thing as ‘just drinking the wine’ (would it be anything like our conception of wine if
it was something we just drank, without speaking about it in any way?), or ‘just listening to a
poem’, analogous to the way that we most of the time completely unawares keep breathing
the air around us and keep hearing the ambient hum and rumble of machines and traffic
around us? And without reflexive and normative commentary on the quality of the wine, or
of the air, would we be able efficiently to protect ourselves from poor or even vile wine, and
polluted or toxic air?
I by no means assume that there is uniformity of judgement, among readers or among
stylisticians, as to what constitute ‘agitated sentences’ or ‘calm sentences’, any more than
8
there is uniformity of view among language-users as to what is correct or faulty grammar or
pronunciation. On the contrary there is divergence, claim and counter-claim, and an ongoing
conversation: this is part of what makes these questions sites both of normativity and cultural
interest. In stylistics I have always argued that the practitioner is aiming, by the rhetoric of
their deployment of linguistically-formulated evidence, to persuade others that this matrix of
formal features, in this assessment of the entextualised situation, is likely to have created in
the reader and likely to create in future readers such and such effects (meaning, interpretation,
signification) on the balance of probabilities. Any higher standard, such as certainty beyond
a reasonable doubt, is incongruent with the theoretical assumptions made so far. The
stylistician is always making a (civil) case, not delivering a proof; they are always making a
normative claim, if you like, and norms change thank God, so stylistic analyses of
Shakespeare will never conclude, final solutions as to his sonnets’ meanings will never be
reached (this clearly creates a tension with iconicity’s claims of natural and enduring form-
meaning ‘fit’). They are also always functioning as a first-person reporter on their own
linguistic experience of the poem--who else’s could they truly reliably report?—or of others’
reported experiences of the poem.
By the same token, if I accept that a redefined linguistics must recognize that how
language-users make language significant (sc., how they attribute or impose form upon its
substance) is contingent upon the individuality of person and situation (in Taylor’s words:
1990: 143), nevertheless the wish of the stylistician is to believe that for this poem or that
play, the individualities of person and situation on the production or the reception side are
attenuated or transcended. Thus the stylistician’s wish is that the form of literary works and
the addressee experiences they provide were comparatively permanent or stable, albeit within
a world where all linguistic form is situation-dependent and variable. Nor, another
qualification, is the stylistician entitled to assume that the literary critical assessments
(menacing, agitated, calm, dynamic), invoked as the reader responses for which the linguistic
commentaries purport to be a causal explanation, are themselves stable, transparent, and
determinate in meaning. On the contrary, they too are inevitably contingent and
impermanent, and some will wax others wane with change of time and place, in terms of
prominence in the literary stylistic conversation: consider the changing fortunes of critical
terms such as racist, gendered, sentimental, muscular, tone-deaf.
One of the things integrationism dislikes about the practices of stylisticians are those
points in their books and articles where they claim to explain the meaning of some form, as if
the identification of the form and of the meaning and indeed the explanation were not just
good for the writer themselves but for readers in general, and for all time. Of course
stylisticians’ explanations, given integrationism’s situationism, cannot be of such an absolute
kind. But it is worth noting how different are the terms of engagement when someone
attempts a stylistic analysis of a Shakespeare sonnet, by comparison with the canonical
situation of linguistic communication within which most integrationist examples are
discussed, highlighting the situation-embedded coordination of meanings which might be
called the normal practice. In two important ways, the sonnet-analysis is different. First,
while the substance to be interpreted is writing, it is mostly writing that we understand is a
script for what actually needs to be spoken, performed (all the attention to rhyme, assonance
and intonation would make no sense otherwise): it is hard to think of another culturally-
entrenched linguistic activity which hovers quite so ambivalently between the categories of
speech and writing. Second, it is understood to be composed by one but intended to be read
by many; and the ‘one’ here is not any old one, but the greatest of verbal artists.
4. The importance of literature, iconicity, and creative ‘fit’
9
I want to return to the ‘Is nothing sacred?’ rhetorical question, which I mentioned earlier in
relation to Nigel Love’s 1985 commentary on those language games that are permitted, and
those that are not, within the fixed-code paradigm. It is a question of enormous resonance. It
was invoked, with obvious irony, by Salman Rushdie also in the 1980s, when some of the
wordings in his Satanic Verses were deemed by some to be so wicked that a death-sentence
was passed on him. But in the predominantly secular Britain of the late 20th
century, the
years of Derrida and Dawkin but really a zeitgeist that reached back to Darwin and the 19th
century rise of scientific explanation, the prospect arose, and is embraced by integrationism,
that nothing linguistic is sacred. More specifically, nothing verbal or linguistic is, by virtue
of its specific wording in such and such a context of situation, guaranteed to signify in some
particular predetermined way. No meaning, whether deemed banal and everyday or
transcendent and transformative, can be predeterminately assigned to some specifiable extent
of linguistic form, and guaranteed as good over time and across a whole speech community:
determinations as to form and meaning are ratified by language-using language-making
participants in the course of the communicational event in which they are engaged. There are
ways in which the briefest reflections on our own changing lives, and on the changes we see
in language through time and space, remind us that we know this. We know that forms and
meanings are constantly in process of revision, renewal, ratification or sanction, promotion or
demotion. But we also yearn for occasions of form-meaning permanence and exceptional
power, transcendent language, celebrated communally as if available to all—usually written
occasions or at least scripted-and-performed occasions. Many once looked for these
occasions in religious texts, tellingly called ‘scripture’ in the Christian tradition, and plenty
still do. Some look instead or as well to other form-meaning communicational pairings, in
music, dance, film, and the pictorial and plastic arts. Some look to literature. And iconicity
discourse reflects that desire.
The discourse of iconicity has a crucial place in the story modern stylistics tells about
iself, alongside such ideas as foregrounding, deviation/markedness, and parallelism or
repetition. It is strikingly difficult to find an entirely satisfactory single defninition of what
iconicity amounts to, for stylistic purposes (the work of Fischer, Nänny and others is
frequently cited, however). In one set of terms or another it re-asserts a natural or logical
connection between the form and meaning of signs, whereas modern linguistics has declared
the connection to be purely arbitrary and conventional. So terms like analogy, similarity, or
resemblance are invoked: the linguistic form is said to reflect, match, enact, be mimetic of or
proportional with, the meaning. Sometimes the match is claimed to be between the form and
the (real) world. And for some (e.g., Jeffries 2010) the iconic text must have an an
‘experiential’ effect: you don’t just read or process the iconic language in poems, you
experience it, the experience being said to ‘fit’ the meaning, the effect that communities of
readers agree a poem has on them. Stylisticians say there is iconicity, or that the language is
iconic, where they can see and demonstrate a degree of calculated ‘fit’ between the sounds or
written shapes used, or the sequences of words used, or the grammar of a whole line or
sentence—in short, the form--and the meaning we believe the poet or writer wanted at that
point to convey, i.e., wanted the reader to experience. Like other discourse analysts,
stylisticians here lean heavily on paraphrase: having sought acceptance of their paraphrase of
the text’s meaning, they argue that the poet’s chosen forms achieve that meaning/experience
entirely more fitly than the paraphrase.
A famous poem involving ears of wheat may serve as an example: Seamus Heaney’s
‘The Harvest Bow’, often read as an homage to his father. Here is its first stanza:
As you plaited the harvest bow
You implicated the mellowed silence in you
10
In wheat that does not rust
But brightens as it tightens twist by twist
Into a knowable corona,
A throwaway love-knot of straw.
The stylistician’s iconicity question is roughly as follows: As we read Heaney’s poem about
the harvest bow, the delicate decorative ‘knot’ that is tied out of a ripe bunch of wheat ears, is
there anything curved, wound, or knotted back on itself, in the language or form (including
the sounds), or in the grammar, that seems uncannily to fit the topic especially well? Is the
poem itself, as an expression of what it is about, amount to what Philip Larkin in another
context called a ‘frail travelling coincidence’ (overtly describing a train journey, but covertly
encapsulating his poem about the train journey: second-order morphing to become first-
order)? Some stylisticians, some of the time, would try to persuade readers that there are
iconicities of form, and will care a great deal more about it than about you was and I were,
judgement and judgment, connexion and connection, and other kinds of more routine
conformity and nonconformity.
In a recent article (Toolan 2015) I tried to define ‘stylistic iconicity’:
By ‘stylistic iconicity’, I refer to that impression of exceptional ‘embodiment’ of a
sign’s meaning in its materiality, such that the communication seems to be a more
direct and natural enactment of the meaning than the arbitrary and symbolic
signification which we ordinarily look for in language (Fischer, 2014; Leech, 2008:
114–15, 149–50). To claim that language is being used iconically is to claim that, to
some degree, the form embodies the meaning; linguistic iconicity is at work when the
form evokes the meaning because it resembles (visually or aurally) the thing meant.
As such, in (rare) cases of what we might call ‘absolute iconicity’, an addressee
should be able to grasp the meaning without prior knowledge of the language in which
it is carried.
There are plenty of things in the above formulation that need adjusting, to suit integrationist
ears, but I think there is sufficient modalizing in the account to protect it: the references to
impression, claims, and what seems to occur. There is an implication too that iconicity is a
natural form-effect fit, rather than something sometimes arising from conventions and
training; I think the implication broadly correct, but it also true that literary education may
encourage our attending to iconic effects (stylisticians are not the only people to have noticed
them). In any event even stylisticians know that iconicity in poetry is rarely if ever an
embodiment, in the linguistic form, of the meaning (the ‘embodiment’ metaphor reflects the
same yearning for ‘real’ and certain intersubjective material contact, the thing itself and not
just a sign or trace of the thing). More typically iconicity is postulated where something in
the form seems consciously chosen (by the poet using her superfine metalinguistic
awareness) for the way it supports or enables meanings that are mostly assessed by other
integrational means. And Leech and Short’s (2007: 195) warning against using iconicity
claims to foist ‘private and whimsical responses’ on literary texts certainly needs to be
attended to; still, it is the same Leech (1969, 2008) who, over many years, has recognised the
importance of iconicity to poetics and stylistics. “How language works ultimately cannot be
separated from [among other things] how its users think it works”, Love wrote in 1989 (817);
and one of the things poetry-lovers and stylisticians believe (yes, it is a conviction) is that
they may be justified in claiming that a poem’s sound echoes the sense, or that the
appearance on the page, the rhythm, the lineation, or the grammar are iconically matched to
the sense, the effect as experienced by the reader.
I have argued (Toolan 2015) that literary creativity (e.g., in a poem), while not
demonstrably different in kind from language creativity in non-literary situations, is
nevertheless distinctive because it is is constituted by the text’s exceptional fitness in
11
addressing some important purpose or situation of which the reader was not aware until they
encountered it in the poem itself. Of course the particulars of a purpose or situation described
in a poem will be unfamiliar and ‘new’; here I am referring to the general contours of the
poem’s purpose, as themselves new to the reader. A new situation is matched with a new
‘solution’, and the fitness of the latter to the former feels inspired: those, for me, are the
components of literary creativity.
So in answer to my own earlier question—
Is stylistics doomed to one ‘second order’ descriptive system after another, in a futile
effort to capture the essence, the life, of the literary text(s) it analyses?
--my answer is that Yes, Stylistics is bound to fail if, ignoring the time-bound fluidity of
signification, it imagines that a permanent essence of a text and a timelessly stable descriptive
system to apply to it are coherent goals. On the other hand, if it is accepted that in the face of
complex and polysemous literature, discursive commentary remains useful which draws on
linguistic terms and principles and aims to further an understanding of how and why a text is
composed in one way rather than another, then all talk need not cease. A non-transcendental
stylistics, deploying various second-order linguistic descriptions while remaining alert to
integrational semiological axioms and the Circean attractions of the language myth, continues
to offer valid and useful commentary on how readers read and how writers write. In its very
provisionality it is arguably a more sophisticated sub-discipline of linguistics than some other
more prestigious ones.
Acknowledgement: I am most grateful to Rowan Mackay, Bettina Perregaard, and Talbot
Taylor for their invaluable comments on an earlier version of this paper.
References
Fischer, Olga and Max Nänny. 2001. "Introduction: Iconicity and Nature." In Fischer and
Nänny (eds) Iconicity. Special Number of the European Journal of English Studies
(EJES) 5, 1: 3-16.
Fischer, Olga and Max Nänny (eds). 2001. The Motivated Sign. Iconicity in Language and
Literature 2. Amsterdam: Benjamins.
Gavins, J. 2007. Text World Theory: An Introduction. Edinburgh: Edinburgh University
Press.
Jeffries, Lesley. 2010. ‘The Unprofessionals’: Syntactic Iconicity and Reader Interpretation
in Contemporary Poems. In D. McIntyre & B. Busse, eds., Language and Style,
London: Palgrave, 95-115.
Leech, G. N. 1969. A Linguistic Guide to English Poetry. London: Longman.
Leech, G. N. 2008. Language in Literature: Style and Foregrounding. London: Longman.
Leech, G. N. and M. H. Short. 2007. Style in Fiction. 2nd
edition. London: Longman.
Love, N. 1989. Transcending Saussure. Poetics Today, 10:4, 793-818.
Love, N. 1990. The locus of languages in a redefined linguistics. In: Davis, H.G., Taylor, T.J.
(Eds.), Redefining Linguistics. Routledge, London, pp. 53–117.
Love, N. 2004. Cognition and the language myth. Language Sciences 26 (6), 525–544.
Nänny, Max and Olga Fischer (eds) 1999. Form Miming Meaning: Iconicity in Language and
Literature. Amsterdam: Benjamins.
Taylor, T. J. 1990. Normativity and linguistic form. In: Davis, H.G., Taylor, T.J. (Eds.),
Redefining Linguistics. Routledge, London, pp. 118-148.
Teubert, Wolfgang. 2010. Meaning, Discourse and Society. Cambridge: CUP. Toolan, M. 2014. The theory and philosophy of stylistics. In P. Stockwell and S. Whiteley,
eds., The Cambridge Handbook of Stylistics, Cambridge: CUP, 13-31.
12
Toolan, M. 2015. Poetry and poetics. In R. Jones (ed.), The Routledge Handbook of
Language and Creativity, London: Routledge, pp. 246-261.
Werth, P. 1999. Text Worlds: Representing Conceptual Space in Discourse. London:
Longman.
Recommended